Observed and Modeled Tsunami Currents on California's North Coast

In 2009, a pilot project was implemented in Humboldt Bay, near Eureka, California to measure the currents produced by tsunamis. This area is susceptible to both near- and far-field tsunamis and has a historic record of damaging events. Crescent City Harbor, located about 100km north of Humboldt Bay, suffered 20 million in damages from strong currents produced by the 2006 Kuril Islands tsunami and an additional 16 million from the 2011 Tohoku-oki (Japan) tsunami. We deployed a Nortek Aquadopp 600kHz 2D Acoustic Doppler Current Profiler (ADCP) with a one-minute sampling interval in Humboldt Bay, near the NOAA tide gauge site. The instrument recorded the tsunami produced by the Mw 8.8 Chilean earthquake on February 27, 2010 as well as the Mw 9.0 Japanese earthquake on March 11, 2011. Currents from the 2010 tsunami persisted in Humboldt Bay for at least 30hrs with a peak current amplitude of 0.3m/s. The 2011 tsunami signal lasted for over 86hrs with a peak amplitude of 1.2m/s. Strongest currents corresponded to the maximum change in water level as recorded on the NOAA tide gauge, about 90min after the initial wave arrival. Tsunami currents associated with ebb tides (tidal currents flowing out of the bay) were about 25% larger than currents associated with flood tides. No damage was observed in Humboldt Bay for either event; the 2011 tsunami pulled one boat away from its moorings at the marina about six kilometers away from the instrument site. Although we have no instrument in Crescent City, we were able to estimate currents for the first three and a half hours of the Japan tsunami using security camera video footage from the Harbor Master building across from the entrance to the boat basin, about 70m away from the NOAA tide gauge site. Most of the damage occurred within this time window. The strongest currents reached 4.5m/s and six cycles exceeded 4m/s in the three and a half hours of data. We used the MOST (Method of Splitting Tsunamis) model to compare measured currents to numerical predictions. MOST does a reasonably good job of predicting peak amplitudes for the 2010 and 2011 events in Humboldt Bay and the 2011 tsunami in Crescent City. For Humboldt Bay, the model does a good job of replicating the first four hours of the signal although the ebb currents are slightly underestimated. The model predictions break down for the later part of the signal. This project shows that ADCPs can effectively record tsunami currents for small to moderate events. Data from this project will be used to validate and/or calibrate MOST so that realistic tsunami current hazard maps can be generated for California for use by harbor managers.